Dazzled to invest

(1)

Av: Jessica Österberg

Handledare: Cheick Wagué Examinator: Maria Smolander

Södertörns högskola | Institutionen för Företagsekonomi Kandidatuppsats 15 hp

Företagsekonomi C | Vårterminen 2016 (Internationella Ekonomprogrammet)

- Are the funding masses falling

for effects?

(2)

Preface

"Education is not the filling of a pail, but the lighting of a fire."

- William Butler Yeats (1865-1939)

_______________________ Jessica Österberg

(3)

Abstract

The issue investigated is the nature of reward-based crowdfunding being too

influenced by visual media and the creator’s personalities rather than focusing on the creator’s ability to run a business and the business plan. Through Source Credibility Theory three sets of categories for the studied variables were developed in addition to one category concerning external actors. The variables aim to measure the visual elements, the honesty and personal credibility of the creator as well as the business plan. A sample of 150 campaigns was picked from the platform Kickstarter, which consisted of half successful and half failed campaigns. Each campaign was observed individually and the communicated elements were recorded and compiled for

analysis. The variables were recorded as binary or continuous and were tested through chi-square, regression and correlation tests. The study concludes that some business plan variables are often communicated for both failed and successful projects, however rarely in detail. The campaigns are often utilising visual elements but the honesty and personal credibility of the creator is not as common. Generally the successful campaigns are communicating all variables to a greater extent than the failed projects.

(4)

List of Figures and Tables

Figures

Figure 1, Agent/Principal Relationship 10

Figure 2, Information asymmetry 10

Figure 3, Variable Categories 20

Figure 4, Quality Video 21

Figure 5, Not Quality Video 21

Figure 6, Quality Picture 22

Figure 7, Not Quality Picture 22

Figure 8, Quirky Element. 23

Figure 9, Chi-square Formula 25

Figure 10, Regression Formula 25

Figure 11, Correlation Formula 26

Figure 12, Scatterplot Numbers of Pictures 27

Figure 13, Campaign Origin 30

Figure 14, All Variables, Yes/Quality/Mentioned 31

Figure 15, Expertise Category 32

Figure 16, Trustworthiness Category 32

Figure 17, Visual Category 33

Figure 18, Correlation Matrix 36

Tables

Table 1, Criteria Video Quality 21

Table 2, Criteria Picture Quality 22

Table 3, Coding 24

Table 4, Regression with Binary Variables 26

Table 5, Regression Models 28

Table 6, Regression Models with Equations 28

Table 7, Most Common Variables 31

Table 8, Variables with Biggest Difference 31

Table 9, Chi-square Differing Variables 33

Table 10, Combinations of Variables 34

Table 11, Number of Backers for Combinations (Yes/Quality/Mentioned) 34 Table 12, Number of Backers for Combinations(No/Not Quality/Not Mentioned) 35

Table 13, Chi-square Test for Combinations 35

Table 14, Descriptive Statistics 35

Table 15 Regression Results 37

Table 16, Regression for “comments”, “picture quality”, “risks” 37

(6)

Introduction

Background

“Well, I think that there's a very thin dividing line between success and failure. And I think if you start a business without financial backing, you're likely to go the wrong side of that dividing line.” -Richard Branson (Ted Talks 2007)

The costs of a business

Starting a business does not only entail coming up with a new and exciting concept or product. To start and run a business resources are required, all resources no matter if they concern the developing of a product, marketing, the business model, rent or staff will come at a cost (Carmela 2016). To cover these costs and make this business possible, financing of some sort is needed. All businesses no matter the size can use different forms of financing, internal or external. Internal financing concerns the company’s equity. For a stable company this could be an option, however for new or small businesses it is not (Fallon Taylor 2015; Calacanis 2011; Griffith 2014). External Capital

External capital comes in different forms, large corporations can utilise loans to a greater extent than small businesses and start-ups. The larger corporations have a more attractive position seen from the banks’ perspective, than the unstable start-ups and small businesses. However the accessibility to bank loans do not correlate with the need for resources. More creativity and riskiness are demanded of Start-ups and small businesses to achieve the required financing, capital is raised through their own funds or by the help of friends and family (Fallon Taylor 2015).

However most businesses require more capital than people can save themselves or lend from relatives. This amount of capital demands turning to banks and apply for a business loan. The banks on the their part are not known to be overly generous with their credit, the banks will thoroughly scrutinise the business plan assessing, for instance, its riskiness and the experience of the entrepreneurs (Hakenes 2004:2412; Sagner 2011:39). Therefore a lot of start-ups won’t receive funding and in extension never get the chance of developing their business idea.

Except for business loans from a bank there are business angels that can offer capital and expertise, but to receive the mercy of a business angel contacts are required and their help is also limited. The business angels and venture capitalists also have criteria in which they decide who or what to invest in, where they target high returns (Hillier et al. 2013:544; Gerrit et al. 2015:2; Rao 2013).

Crowdfunding and its history

Crowdfunding has emerged as an alternative informal financial market for start-ups, small businesses and artists. Inspired by crowdsourcing, a form of outsourcing, where the tasks are announced on the Internet for anyone who wishes to contribute and have the resources can do so. For crowdfunding, instead of using other people’s

(7)

crowdfunding platform, describing their project or business, what they want to do and how they’re going to do it. Anyone can then decide to give them a small contribution to achieve their project and in return they get a reward or equity in the future

company.

Through crowdfunding start-ups and other projects, and in some cases ordinary people in need of money for ordinary things can receive funding. Small contributions added together result in the creator reaching the financing goal. Instead of being granted a business loan for the entire sum of money needed or having a venture capitalist invest, entrepreneurs can post their project or business idea on the Internet and have the finances raised by anyone who believes in the project. The inventiveness lies in many small contributions adding up to new innovation. Since crowdfunding entered the financial market roughly a decade ago the phenomenon has developed and grown substantially. The number of crowdfunding platforms is projected to reach over 2000 in 2016 (Drake 2016). These platforms operate on the global market of the Internet, competition is fierce and most projects launched won’t receive the funding they aim to (Mollick, 2014:4,13).

The growth and development of crowdfunding has made the users of this financial market reach new innovative dimensions, and there are now even crowdfunding consultant agencies that can help creators with their campaigns

(Crowdfundingstrategy.net 2014), turning the actual campaign creation segment of the process into its own market. In 2012 the aggregated funds raised on crowdfunding platforms was 2.7 billion dollars, which by 2013 increased, reaching 5.1 billion dollars (Barnett, 2014b). The concept has also been used for internal marketing at large corporations to give employees the opportunity to create new innovations and choose which should be funded (Muller et al. 2014:510).

Different forms of CF

There are four different forms of crowdfunding; the charitable form where backers donate money without anything but altruistic reassurance in return. As for the loan-based the backers offer loans to the creators, where the backers can decide what interest they want and when this interest should be paid. The two most common forms of crowdfunding are equity-based and reward-based. For the equity-based the

investors buy shares in a future company, the motivation is the future returns on these future shares. This particular sector of the crowdfunding industry is characterised by funders being informed to a higher degree and as a result also contributing with more significant sums (Gerrit et al. 2015:3) In difference the reward-based crowdfunding projects are supported by a larger number of backers contributing smaller amounts of money and are in return given a product or service (Mollick 2014:3).

Dimensions of CF

The campaigns that succeed in their funding goals do so thanks to the backers believing in their business. The creators now have the pressure of the backers to deliver the rewards or equity that was promised in return for their backing. Not all projects will deliver according to plan and the projects that receive more funding than they set out to get are more likely to suffer great delivery delays (Mollick, 2014:12). And some projects won’t deliver at all; some simply fail their projects due to lack of contacts, contracts and incompetence, others are used to deceit through fraud. The crowdfunding platforms are not subject to any regulation; the only requirements to

(8)

post a project are dictated by the platform organisations themselves. These

organisations have an incentive to create barriers to keep the projects listed on the website sincere. The creators themselves then need to convince the public to fund their project. The informal financial market of crowdfunding is highly dependent on credibility and trust, the platforms need to appear credible, and the creators need to win trust to receive funding.

Problem

There are many articles and websites giving recommendations on how to successfully launch and execute a crowdfunding campaign, most of these recommendations

concern the marketing of the campaign, the importance of telling a story, making a video and of social media (Barnett 2014a). Little of the advice relate to developing a convincing business plan or the product itself.

Earlier research also supports the affect that a video has on the campaign’s success. Stating the importance of putting effort into the campaign, creating quality signals that are perceived as credible and therefore make the campaign more likely to receive funding. The element of social media is also raised as a factor that could be valuable for success (Mollick 2014:13). With this outline, according to experts and research, targeting visual effects and being likeable will make your crowdfunding campaign more likely to succeed.

The banks and business angels are professional decision makers when it comes to grant loans and invest in new businesses. The public who invest in crowdfunding projects, however, are amateurs in the field of deciding whether a business is worthy of investments or not. Not having the same experience and know-how makes the backers easily misguided (Gerrit et al. 2015:2). Earlier research concludes that some funders participate to expand their social network and are also motivated by the participating factor itself (Greber and Hui 2013).

Where a bank will scrutinise your business plan, experience and the project’s

riskiness and a business angel or venture capitalist will consider if the idea is scalable and give high returns (Gerrit et al. 2015:2; Hakenes 2004:2412; Sagner 2011:39), the masses are more attracted to softer values, for instance, does the crowdfunding campaign tell an interesting story? Corporations are also using crowdfunding as a marketing opportunity, both externally and internally (MacDougal 2014). In a way stealing the attention from other projects, making the small businesses compete with the resources of large corporations. This could leave innovative projects that lack these certain marketing skills without funding.

There’s also a certain spectacle to this new market, celebrities like Zac Braff and James Franco and the TV-series Veronica Mars have benefited millions of dollars from the funding masses through their crowdfunding campaigns to create different creative projects (Kickstarter 2014a,b; Indiegogo 2014).

Is the nature of crowdfunding now characterised by projects and creators being likeable, rather than having good business ideas? Has crowdfunding turned into a

(9)

financial market that’s more a marketing competition than a way for new innovative ideas to become profitable businesses?

The crowdfunding phenomenon emerged as an instrument for start-ups to get funding through the public. If the focus has gone from innovation and business ideas and products to a good marketing campaign the crowdfunding market is in trouble. They risk producing projects that only know how to perform good marketing campaigns which isn’t desirable unless they are aiming to start a marketing agency, the whole concept of crowdfunding would be lost. If this development is the reality, this new informal financial market is in danger of the inefficiency of the masses being attracted to flair.

Purpose

The purpose of this study is to investigate what kind of elements affect the success of a reward-based crowdfunding campaign, and also if these differ from the campaigns that don’t succeed.

Research Questions

I. Are campaigns that don’t communicate their business plan successful in reaching their funding goal?

II. Are campaigns that communicate visual and personal credibility elements successful in reaching their funding goal?

III. Are there any differences in the nature of the communicated elements in successful and failed campaigns?

IV. Are there any combinations of variables that are more commonly used by the successful campaigns?

Limitations

The population for this study is based crowdfunding projects. The reward-based crowdfunding is through its small contributions targeting everyday people; it’s the characteristics of this group's decision-making process that are relevant for this study. Since the study investigates start-ups and businesses, the study will be limited to campaigns launched by businesses and not one-off projects. There are no

geographic limitations; the nature of crowdfunding has no borders since it’s based on the Internet.

(10)

Theoretical Framework

In this section the theoretical framework including theories and earlier research are presented and discussed. Agency Theory, Behavioural Finance and Source Credibility Theory together build the framework for the study. The main concepts of the theories are accounted for as well as illustrating the aspects that are relevant for crowdfunding and this study. Earlier research follows where other studies that relates to this research are disclosed.

Theories

1..1 Agency theory

One major aspect of the agency theory concerns the difference in the agent and principal’s goals. One of the theory’s main arguments lies in the issue of information asymmetry; the agent and principal have different sets of information (Akerlof 1970:490; ). The key for this theory lies in the agent working on behalf of the principal, in finance and management theory the agent is normally the board of a corporation and the principals are the shareholders (Ross 1973:134; Mitnick 1975:27). Other agent/principal-relationships are between an insurance company and a

corporation or a person, between the bank and a corporation or person or lastly between a funder and creator in a crowdfunding campaign (Shavell 1979:66).

The agent is running the corporation and is involved in its day-to-day operation and therefore has a unique insight into how the corporation is performing. However the principal who is not involved in the management of the corporation is dependent on the information given by the agent (Akerlof 1970:489-491). This asymmetry in information leads to complications in the relationship between the agent and

principal, it is an issue of risk, if the principal has insight in the agent’s work the risk will be perceived as smaller (Shavell 1979:56).

Figure 1, Agent/Principal Relationship

(11)

In regards of asymmetric information there are two main issues, moral hazard and adverse selection. Asymmetric distribution of information results in important decisions being made on incomplete information; this means that asymmetric information is one factor that increases the risk for the parties involved.

Adverse selection can be described as the better-informed part, the agent has an incentive to exploit the principal, displaying an imbalance of power. The different parties develop strategies to address the issue of power imbalance to lower the risk, screening and signalling. (Garbade & Silber 1981:501; )

Signalling

The main concept of signalling is that agents can, through actions, send signals to the principal. When an agent executes certain activities to enhance the principal’s insight in the corporation, agency theory refers to this as signalling. The action that the agent performs sends signals about the corporation’s condition to the principal (Garbade & Silber 1981:499-500; Casu et al. 2015:10). For example when the agent makes public announcements about the corporation’s new operations, or publishes a policy of being a more transparent business. The previous actions mentioned are to decrease

asymmetric information, however there are other actions that can be practiced to increase the asymmetry or just by choosing not to attempt to decrease the asymmetry is an act to increase it.

Screening

The principals also have the ability to decrease information asymmetry by enforcing different forms of screening. Screening consists of all the actions that a principal performs to scrutinise the agent’s management of the corporation. It’s a form of risk management where the principal try to establish an idea of how the corporation is managed and what issues the corporations have and what impact these factors have on the principal’s interests. This aspect also inheres the results of the screening such as risk premiums, interest rates and dividends. If the principals after the screening suspects that there are uncertainties or that the agent is engaging in risky behaviour the principal will react with higher costs for the agent to compensate (Holmström 1979:74-75; Ross 1973:138; Casu et al. 2015:10-11)

Moral hazard

Moral hazard depicts the risk of an agent acting against the interest of the principal attempts favour its own interests or to do something else than was said in the contract. The latter is mostly of relevance when a bank grants a business loan for a company and the company instead of going through with the idea they were granted the loan for, carry out another riskier project (Holmström 1974:74; Casu et al. 2015:11). Agency Theory and Crowdfunding

In the case of crowdfunding the agent would be the creator and the principals the funders who not unlike a corporation gives the creator money and receives something in return. The difference for reward-based crowdfunding, in relation to other

agent/principal relationships, lies in the return; the principal in crowdfunding will receive a product or service only once and therefore are only concerned with the delivery of this reward and not like the shareholder of a corporation who is concerned about sustainable growth and dividends. (Hillier et al. 2013:37)

(12)

The signaller is the creator of the campaign; they are through what is published on the campaign signalling what the project entails. Since the only communication that the creator has with its potential funders is through the campaign site, anything that is published on the site is equal to a signal. Per definition these signals are the

presentation of the idea itself, the business plan, the timeframe, the video, the rewards and also the dimension of how these elements are communicated. If there are

inconsistency in the manner of which these signals are communicated, such as spelling errors, insufficient business plan or risk analysis or a video that isn’t in a good condition the potential backers might interpret this as an indicator that the project also lack quality. As for quality signals such as a compelling story, high quality video and a developed timeframe the potential backer will assume the project posses the same quality. Earlier research has shown that quality and uncertainty mitigating signals increase chances of succeeding (Gerrit et al. 2015:22) Screening in Crowdfunding

The screening aspect of crowdfunding involves the process where the funders assess the campaigns on the platforms, deciding whether they’re worth the investment. Moreover concerning the relationship between risk and return for funding a project, for the reward-based crowdfunding the return for risk taken is the reward. The funder will often have several different rewards to choose from, the more they spend the “better” the reward. In this way the funders themselves can decide what level of risk they want to be exposed to.

Moral Hazard in Crowdfunding

As for moral hazard within this new financial market, creators might not do what is stated in the campaign that helped them raise funds. There is the risk that creators never meant to go through with the project at all and use the platform for fraud, by knowingly misleading people into backing a project that was never meant to exist. Since moral hazard is a concept that occurs after the actual backing is completed, it won’t be taken into consideration for this study.

1..2 Behavioural Finance

Behavioural finance developed as a reaction to neo classical economic theories’ descriptions on the rational decision-maker. The new paradigm challenged the behavioural assumptions of the “homo oeconomicus”, the foundation of economic modelling and analysis, and a rational human being optimizing all decisions always arriving at the most efficient and cost-effective solution (Fromlet, 2001:64; Simon, 1957:23). Financial decision-making and decision-making was discovered to suffer from irrational behaviour; people lack the ability to make completely rational decisions. Moreover the speed, amount and level of complexity that the information in today’s society contains makes it impossible for people to select, rank and optimise to arrive at the best decision or solution (Fromlet, 2001:65).

There are a number of different aspects of behavioural finance that are applicable for this study is,

I. Heuristics, selective interpretation of information, II. Psychology of sending and receiving messages, and III. Varying availability of information

(13)

As for heuristics the issue of selectively interpreting information, meaning that to make decisions people tend to rely on experience, what kind of decisions in similar situations has proven to give satisfying results? There’s also the development of decision patterns to approach information and decisions in a more practical way (Fromlet, 2001:65).

Actors on the market are also incorporated since they can decide to communicate messages in different ways; Fromlet (2001) illustrates this through comparing two newspaper headlines that have the same message communicated in different ways, one more pessimistic the other optimistic. Depending on what is communicated, people will interpret the information in regard to how it’s portrayed; this describes the psychology of sending and receiving messages (Fromlet, 2001:66).

Varying availability of information names the issue of information not being available for everyone, financial markets are not actually perfect. In addition to information inaccessibility there is also the matter of having the skills to interpret and understand the available information (Fromlet, 2001:66).

Satisficing

Satisficing is a concept, first described by Herbert Simon in 1947, where a person making a decision not actually seek the rational and most optimal solution but a sufficient solution. To find the most optimal decision is not only time-consuming but also many times impossible, therefore people satisfice (Simon, 1957:25). For

instance when one gets dressed in the morning one doesn’t calculate the most efficient way of getting dressed and what clothes to wear, just putting something on is actually more efficient and most times also sufficient to keep comfortable and respectable. Satisficing is also relevant for behavioural finance where participants in the

marketplace find the most sufficient solution for the task at hand, since there’s often a lack of resources and time for optimising, maximise the most beneficial option. Behavioural Finance and Crowdfunding

Behavioural finance offers guidelines to how people make economic decisions, the limits in rationality and the element of satisficing. The theory has relevance in the case of this study since it could explain how funders make decisions in terms of satisficing, how messages are received, and how they interpret, choose and understand information.

1..3 Source Credibility Theory (SCT)

Source credibility illustrates how trust and confidence in a person or entity that is sending messages, through actions or writing, is perceived by the receiver.

There are four contexts a) presumed credibility, b) reputed credibility, c) experienced credibility, and d) surface credibility.

Presumed credibility is when someone believes that something is credible based on earlier assumptions. When someone believes something is trustworthy because a friend has ensured him or her; this is reputed credibility. Experienced credibility entails the credibility for when someone perceives something as credible due to having experience of it. The kind of credibility that is of most relevance for this study is surface credibility (sometimes called visceral credibility), which focuses on

(14)

et al. 2014:65-66).

The theory puts its main focus on the receiver’s perception rather than the characteristics of the source (Simpson and Kahler, 1980:18). There are three dimensions concerning the theory

I. Trustworthiness, II. Expertness, and III. Dynamism

The trustworthiness dimension is described by the perceived integrity and honesty of the source. Hovland and Weiss (1951) come to the conclusion that the communicator has a direct influence over the perceived credibility of the message communicated; trustworthiness in the source affects the opinion of the receivers. Expertise represent how experienced, competent and authoritative the source is perceived to be, lastly dynamism expresses in what manner the message is delivered; in spoken

communication for instance this could be described as charisma. In written communication how the message is presented is a more accurate illustration,

dynamism could be associated with a message being presented in a bold and energetic tone (Lowry et al. 2014:66).

SCT Online

Source credibility theory in online environments have previously been subject for study, Robins and Holmes (2008) performed a study for what influence website aesthetics had on credibility. The study as conceptualised through two categories; Low aesthetics treatment (LAT) and high aesthetics treatment (HAT), where the former represent the websites that lacked professional design and planning and for the latter the websites that were planned strategically and professionally through design, colour and layout to fit the brand (Robins and Holmes, 2008:387)

They concluded that in 90% of the cases treated aesthetics did increase credibility, proving that the visceral credibility and dynamism did succeed in capture consumer, adding importance to the first impression when visceral credibility is formed. The initial hypothesis that treated aesthetics has an interaction with perceived credibility, a website with compelling aesthetics will be perceived by the receiver as more credible that one that doesn’t. Meaning that a business that wants to be perceived as credible has to put thought in the surface credibility aspects and leverage dynamism elements (Robins and Holmes, 2008:390, 398).

Source Credibility Theory and Crowdfunding

SCT applied to crowdfunding results in the creator being the source and the backer is the receiver. Robins and Holmes’ (2008) conclusions regarding the visceral

impression can be used for this research since there are similar elements to a business’ website and a creator’s campaign, both target consumers and rely on an Internet forum to capture their attention and trust. Dynamism, how the message is delivered, is crucial for these kinds of actors to achieve their goals, in the crowdfunding aspect this means how the campaign site is designed and planned.

There is also relevance for the cognitive decision-making processes like

(15)

dimension of expertise is also applicable, if the creator shows signs of being experienced and competent.

Earlier Research

1..1 Dynamics of Crowdfunding

In 2014 Ethan Mollick performed an exploratory study of crowdfunding, the aim for his study was to procure a general understanding of the phenomenon as a whole. The study investigated what projects were successful and why, and if they were did they actually deliver. Does funders decide to fund because the project signals quality or are there other less rational reasons behind the successful projects?

Projects of all categories were studied from a range of different variables including comments, number of funders, different quality variables and capital goal.

Mollick (2014) came to the conclusion that projects that signalled quality were more likely to be successful, funders to some extent made rational investing decisions not unlike a professional.

Quality was measured by three criteria a) if the creator published a video b) if the creator posted updates after campaign launch c) spelling errors in the campaign. Through these three criteria Mollick aimed to measure effort and by effort quality, meaning that if a creator put enough effort into the campaign by posting updates, a video and checking the spelling, it signals quality to funders.

Other discoveries made from this study were that most projects actually fail to receive funding, and those projects lucky enough to succeed with a large margin often failed to deliver on time, if at all.

Mollicks research shows that creators are more likely to succeed if they put more effort into their campaign, the campaign most likely won’t be successful, but if it receives funding far beyond the pledged sum they won’t be able to deliver. Relevance for this study

Mollicks study on the dynamics of crowdfunding sets the outlook for continued crowdfunding research. Since this study targets the quality of the signals against how detailed the business plan is, the quality aspect is of greatest relevance since the quality was proven to be an important factor in a campaign succeeding.

1..2 Signaling in an Online Environment

The study investigates the signals undertaken by American e-commerce

pharmaceutical companies. Signals are crucial for the e-stores to sell their products online. The authors state the importance of signals on the website for an e-store since they lack the purchasing process characteristics of a physical shops (Mavlanova et al. 2012:240). The consumers needs to make purchasing decisions based on the

information that the businesses decides to publish on their website. The signals are crucial to bridge the information asymmetry between seller and buyer in the online market.

(16)

certain signals stating their quality as merchants and the quality of their products. The aim is to persuade the consumer that their business is credible and performs high quality operations.

These signals are conceived at different costs, based on this the authors divided the signals into high and low quality. The high quality signals are expensive and in some cases demand a high degree of effort, for instance being granted a certain stamp of quality by an external organisation (Mavlanova et al. 2012:242).

The authors concludes, in accordance with stated hypotheses, that low quality sellers are more prone to use low quality signals at a low cost where high quality sellers aren’t hesitant to invest in high quality signals. Consumers can through exploring the websites for high quality signals decide whether the e-store itself is of high quality (Mavlanova et al. 2012:245).

Relevance for this study

This study gives insight on the matter of how online platform consumers perceive signals. Further depth is offered through the connection between effort and quality and how it affects the signals portrayed. It also concludes a relationship between low quality signals and low quality companies.

1..3 Signaling in Equity Crowdfunding

Gerrit et al (2015) explore equity-based crowdfunding through signaling theory, with the outlook that small investors lack the financial sophistication of professional investors, such as venture capitalists; the signals that the creators send through their campaign are of great importance. A framework to conceptualize effective signaling is established through two groups, venture quality and level of uncertainty. Venture quality is measured through human-, social- and intellectual capital and uncertainty levels through the amount of equity shares offered and financial projections. The study finds that human capital and both variables concerning uncertainty levels are of significant meaning to succeed with an equity-based crowdfunding venture. In difference to other studies the social network variable did not show significance neither did intellectual capital (Gerrit et al. 2015:22). However a developed board structure and detailed information regarding the risks of the venture proved to be meaningful factors to succeed.

Relevance for this study

The study supports that, for equity crowdfunding, details about risks and board structure are important to succeed, variables that mitigate uncertainty for the backers. Although the study doesn’t concern reward-based crowdfunding, it is still relevant for this study offering a framework that could be used to interpret uncertainty.

(17)

Method

In this section the methods used to answer the research questions will be presented. The first part contains a description of the research design, followed by an

explanation of the processes of selecting variables, platform and industries for the study. Continuing with a presentation of how the data collection and analysis of data were performed and lastly the method is scrutinised in method criticism.

Scientific and Theoretical approach

The research is carried out in a positivistic manner to ensure a higher degree of objectivity, meaning that the data in the study is analysed through natural sciences such as statistics. The positivist base his or her knowledge on empirical evidence, as will this study where the research questions will be answered based on empirical findings from the collected data. Furthermore the research is implemented through a deductive framework where confirmed theories determine the base for the study. The concepts in this study are based upon theories and earlier research that are relevant to crowdfunding, credibility and decision-making (Bryman and Bell 2005:26).

Research Design

The study has a cross sectional design where quantitative data are of interest, the study was performed at one specific time, not during a longer period or at two or more separate occasions as in a case study. The aim for the study is to investigate variation and not depth, therefore a large number of cases will be observed. To answer the research questions many cases need to be studied since only one or a few cases won’t provide sufficient amount of data. Collecting data from many cases will also enable a higher degree of generalizability where an in depth study would scrutinise only one or a few cases to examine a specific phenomenon, but would be unable to generalize this beyond the small number of studied cases (Bryman and Bell 2005:65, 85,100).

Sample

The population of this study is start-ups and businesses that launch crowdfunding campaigns through reward-based crowdfunding. On Kickstarter over 300000

campaigns have been launched, there are also thousands of crowdfunding platforms, as well as the number 300000 is statistics from the launch in 2009(Kickstarter 2016b; Drake 2016). These factors make it difficult to estimate the exact size of the

population, but according to The Crowdfunding Center (2016) more than 400 projects are launched each day and there are beyond 12000 active projects daily, however all these are not start-ups or businesses. The sampling method is a combination of stratified- and cluster sampling. A stratified sampling method splits the population in homogenous groups; this has been done when choosing equal numbers of success and failed campaigns. The cluster sampling method aims to split the population in

(18)

sample for this study has been split between three industries (De Veaux et al. 2014:317-319).

The motivation for using three different industries is to avoid that the characteristics of one industry having too large impact on the results, this could create uncertainty if the result was unique for the industry or if it reflected the characteristics of

crowdfunding or the characteristics of that industry in crowdfunding. Furthermore the sample can’t be considered to be a random sample since all the campaigns in the population did not have the same chance of being picked for the sample, however randomness within the stratified, clustered groups have been targeted (De Veaux et al. 2014:316). The campaigns were picked for the sample by using Kickstarters search-function that allows one to filter the results through different criteria. By sorting the search through “end-date”, meaning that the campaigns that ended most recently show up first. The time the campaign ended does not have any connection to other elements that could negatively influence the sample, and therefore the first campaigns that showed up through this criterion were picked for the sample.

The sample consisted of a total of 150 projects, where 75 campaigns were successful and 75 campaigns failed to reach their funding goal. These 150 projects make out a small piece of the rough estimation of the population that reaches over 12000 active projects daily. The 150 projects were selected in three different categories on the Kickstarter website, Design, Food and Technology. To avoid a large number of one-off projects, the more “artistic” industries like publishing, music and art have been avoided. On Kickstarter these industries are often characterised by projects not meant to last, like for instance record an album or publish a book (Kickstarter 2016a). The large number of these kinds of projects will make the data collection awkward and also three industries are suitable for the size of the sample.

The other forms of crowdfunding, like charitable where backers only donate money, lacks a business perspective and the equity-based and peer-to-peer loan crowdfunding are more complex and therefore demands a higher degree of knowledge from the backers. The reward-based crowdfunding offers the opportunity to make a small contribution and something is given in return, either a thank you or a product or service. This makes it possible for anyone to invest in the project; it is the character of this relationship that is scrutinised in this study.

On the Kickstarter platform the creator don’t receive any money if they don’t reach the funding goal by at least 100% (Kickstarter 2016a), this definition of success will also be used in this study.

The funding goal for the campaigns in the sample was at least 10000 American dollars or the equivalent in any other currency.

These relatively high goals are the limit because projects with a lower capital goal will more easily be reached through help from family and friends. For instance a project with a funding goal of 4000 dollars could through contribution of friends and family succeed, if 100 people contribute with 40 dollars each they will reach their goal. This possibility may result in biases, and therefore projects with lower financing goals are eliminated.

(19)

Selection of Platform

Kickstarter is according to crowdfunding.com the second largest crowdfunding platform on the Internet (GoFundMe 2014) and was launched in 2009. Since the launch over 100.000 projects have received funding and they have over 10 million backers that have pledged more than 2 billion dollars, this makes it an established platform that attracts a large number of backers and creators (Kickstarter 2016b). The variety of projects, industries and nationalities on the platform increases the likeliness of a diversified audience. These factors make Kickstarter an appropriate platform for this study, since the aim is to clarify characteristics of the crowdfunding phenomenon for businesses in general and not a specific industry or segment.

Kickstarter won’t delete any projects off their site to increase transparency, one can search for any project launched on Kickstarter (Kickstarter 2016a). Their transparency is of importance to effectively perform this study since it relies on historical data of the campaigns.

Selection of Variables

To accurately measure the issue at hand 19 independent variables of both binary and continuous character have been chosen in addition to the dependent variable success. The 19 different variables have been grouped in different categories.

The dependent variable that will be compared to the independent variables is the success rate; this variable will be recorded in both continuous and binary form. To answer the research questions the rate of success is of outmost importance to record since the affect the different independent variables have on the success rate is the basis of the study. In addition to the success rate, the number of backers will also be recorded as well as the country the campaign originates from, what industry it belongs to, and whether it is a start-up or already a company.

To effectively determine what variables to measure the Source Credibility Theory has been utilised, the three concepts dynamism, trustworthiness and expertise make out the structure of the development of the variables to measure this issue.

As stated in the theory section, dynamism treats the superficial features of the

campaign, this concept will also be referred to as the Visual category to emphasise the variables visual nature. The components concerning honesty and integrity of the creator is the Trustworthiness category. Lastly the expertise category represents the competence of the creators, and will also be referred to as the Business Plan category (Lowry et al. 2014:66).

1..1 External Variables

These three concepts construct the groups; through this framework the variables have been identified, in addition to these variables that the creator of the project control two external variables that could affect the success of the campaign. They are

(20)

considered to be external since other people or entities than the creator themselves control them; these variables are “comments” and “Kickstarter loves”.

On the Kickstarter platform, the Kickstarter management themselves can endorse campaigns by marking them as “Projects We Love” (Kickstarter 2016d). This variable shows support from the platform itself and adds to the perceived

trustworthiness and expertise of the creators, but the creators themselves do not have the power to communicate this and therefore this variable will be recorded but not added to any of the variable categories but make up their own. People can post comments on the campaign site, without the creators controlling that comment or what they say. Like the Kickstarter endorsement comments could add credibility to the creator and the campaign therefore this variable will also be recorded.

1..2 Dynamism/Visual Variables

Since this concept focuses on superficial features the variables that are gathered in this group are of visual nature (Lowry et al. 2014:66). The key condition to decide whether a variable would be suitable for this group is if it can be observed at first glance, making the natural choices for measurement the video and picture posted on the campaign site.

Measuring Quality

Mollick (2014) measured quality through effort, which also will be a guideline for this study. There’s a distinction between different kinds of videos and pictures. To efficiently measure the use of the visual elements, just reporting the existence of a video or picture won’t be sufficient, since the picture and videos are of different quality. Grouping them together creates biases where interesting differences could be discovered.

(21)

between the high and low quality features in the media published on the campaign sites. The motivation for this is to keep the judgement of variables transparent but also to ensure consistency in the data collection process.

In Figure 4 and 5 illustrates the high and low quality. In Figure 4, high quality, the video is filmed in an environment that looks like a workshop or office. In Figure 5, low quality, there’s a clear home environment and the angle of the frame also gives the impression the creator in the video is filming himself without help from someone else holding the camera. Showing that in Figure 4, more effort was put into making the video than in Figure 5. Another sign for limited effort are videos that are made up by a simple slideshow with or without music. One can easily create a slideshow (Bestslideshowcreator 2013), meaning, as much effort isn’t spent on a slideshow than an actual filmed video. Other criteria for judging video quality are if the sound is bad or if there’s no sound. The definition for bad sound for this study relates to videos where you can’t make out what the person in the video is saying or if there are disruptions in the sound.

Table 1, Criteria Video

(22)

Similar methods are used to decide quality of the pictures; high quality pictures have high definition and are clear and have accuracy for the project. As seen in another example from Kickstarter, Figure 6 shows a picture that is clear and Figure 7 shows a muddled picture that doesn’t give a planned impression.

1..3 Trustworthiness

The trustworthiness variables concern the aspects about the campaign that could make the receiver perceive the creator as honest and decent (Lowry et al. 2014:66), for this study this element is conceptualised as the creator’s personal credibility. The

variables that most accurately measure these characteristics of the creators are pictures of the creators themselves, the creator’s story, which basically is a presentation of the creator who they are and what they have done. This concerns information of the creators’ background that isn’t relevant to the campaign itself; therefore storytelling doesn’t entail experience in the field but personal facts.

Updates on the campaign site have been subject for study in earlier research (Mollick 2014:8) and are also of interest for this study. If the creator posts updates it could make the potential backers perceive that the creators show dedication to the project. One variable for this group concern the credibility of external actors, many projects post references to for example magazines where the project has been mentioned, this factor adds to the perceived trustworthiness of the creators.

Research by Lyttle (2001) concludes that humour can affect the persuasion process therefore the last of the trustworthiness variables is called “quirky element” this variable contains elements in the campaign that could be described as “goofy”, personal facts and humour are key factors for this variable. Below, in Figure 8, is an example of this kind of variable, there are pictures of the creators and also of their

Figure 6, Example: Quality Picture. Source: Kickstarter 2016e

Figure 7, Example: Not Quality Picture. Source: Kickstarter 2016f

(23)

1..4 Expertise/ Business Plan

The variables selected to measure expertise, rational elements, are as many as the other two groups combined. The other two groups focus on visual appearance and the creator’s personal credibility where this group targets other tendencies that involve the business plan and creator experience.

These variables rational manner also leave less room for researcher interpretation biases; the variables will simply be noted if they are mentioned in the campaign. This group consists of the rewards, how many different rewards are offered and how many of these rewards offer something tangible respectively intangible. By tangible and intangible the goal is to differentiate between rewards that only offers a thank you or engraving the funders name on a product, website or inventory. The tangible rewards are regarded as tangible if they offer something other than a thank you or something additional to a thank you, like a product, service or an event.

The risks with the project will be documented if they are mentioned and if any strategies to mitigate these risks are mentioned. All campaigns have a pre-structured subheading on the campaign site called “Risks & Challenges”(Kickstarter 2016c) which is a natural way for the creators to inform the backers about the risks and challenges their projects faces. The fact that this is a pre-constructed element on the site that can’t be removed by the creator could affect the number of projects that mention risks. However distinction has been made between creators that mention risks and creators that state that there are risks concerning the project but not saying what they are. In turn the strategy is only recorded if an actual strategy is mentioned not just a promise to inform the backers if any of the risks become reality.

As for timeframe, budget and creator experience, any of these are mentioned in some way in the campaign will be reported. Because of crowdfunding’s informal manner a timeframe and budget is presented in a simple and approximate way. Therefore timeframe and budget are considered to have been communicated if they are

mentioned in the video or through text by offering a rough estimate of delivery dates and where the funds will go.

(24)

Procedure

1..1 Data Collection

To find the finished campaigns searches were made on Kickstarter using their filter for the three industries, but also filtering funding goal with 10.000 dollars as the lowest limit. To remove the biases that any algorithms used by Kickstarter to highlight certain projects the third filter sorted the campaigns by “end date”. The data collection was made through manually scanning finished crowdfunding campaigns on the platform Kickstarter. The dependent and independent variables were recorded and measured through predetermined criteria to make the results reliable and consistent. For each campaign selected as part of the sample, the data was recorded, compiled and analysed using Microsoft Excel.

1..2 Coding and Recording

Some of the data was recorded through binary coding where the variables of quality or no quality, mentioned or not mentioned. “Quality” and “mentioned” were assigned the value of 1 and “not quality” and “not mentioned” were assigned 0. Quirky element was given 1 if there was such an element on the campaign and if not 0, the same arrangement was made for “picture of creator” and “storytelling”. The funding goal and the final funding was recorded and also the percentage of which the goal was funded. This made it possible to compare different funding goals but also currencies. As for the continuous variables like number of pictures, rewards, updates and

comments; the amount of the variable was counted and recorded. The number of backers was also recorded as well as the country where the campaign was launched, if the campaign belonged in design, food or technology and if it was a start-up or

already a company.

Table 3, Coding

1..3 Data Analysis

The data consists of binary and continuous variables that due to their different natures were analysed separately and then combined. General tendencies of the data were analysed and compiled in diagrams and graphs to obtain an understanding for the data. This data concerned the different countries where the campaigns originated from, the campaigns that received the highest degree of success, the most- and least common binary variables and how many campaigns that had only used variables from one of the variables-categories; Visual, Trustworthiness and Expertise.

(25)

1..3.1 Binary variables

The binary variables were analysed all together as well as successful and failed campaigns separately. The percentages were calculated for the separated data. The variables were analysed to determine the most common variables for successful and failed projects, both the percentage and the counts. The variables that differed the most between the successful and failed campaigns were then identified and tested through a chi-square test, although the differences could be observed statistical significance can’t be concluded by only observing the different percentages. A chi-square test can be used to test independence in the relationship between categorical variables, independence; no relationship is always assumed. The test is performed on the difference between observed values and the expected values (De Veaux 2014:709,713). The H0 hypothesis assumes that the variables were

independent and the H1 that the variables were dependent. Meaning that if the H0 hypothesis was rejected there was a relationship between the variables and success. The six variables that differed the most were each tested against the dependent variable success; the p-value given from the chi-squared test was tried at the 5% significance level.

The most common variables for the

successful campaigns were further analysed through regression. The two binary regression models were constructed with two variables that belonged in the same variable-categories, Expertise and Visual. One of these models had a more meaningful correlation coefficient and r-square value. A regression with binary independent variables is interpreted in a different manner than a regression performed solely with continuous variables. Since the independent variables only can take the value of 1 or 0, the value of y needs to be interpreted accordingly.

Figure 9, Chi-square Formula. Source: De Veaux et al. 2014:709

Figure 10, Regression Formula. Source: De Veaux et al. 2014:534

(26)

The 1 for all binary variables represents the existence of said variable and 0 the lack of it. The dependent variable, level of success, will be affected by the existence of the variable, meaning if x=1 the dependent variable will decrease or increase but if x=0 it will result in the regression coefficient also being 0. Meaning if the binary variable is present it will affect the value of y but if it isn’t present only the intercept and/or the other x-variables will determine the value of y. This is illustrated in the table below, when both x-values are 0; y is predicted to the intercept, only when x=1 does y increase (Skrivanek 2009).

1..3.2 Continuous Variables

The continuous variable’s correlation were first analysed to investigate the linear relationship between them and level of success. The correlation can’t conclude causality between two variables but it does tell if two variables have a linear

relationship. Correlation is given as a value between 1 and -1, any value between 0-1 shows a positive relationship and a negative relationship is between -1-0. The closer the value is to 1 or -1 the stronger linear relationship is. A correlation of -7 or 7 is considered to be an adequate value to conclude a relationship (De Veaux et al. 2014:178).

Table 4, Example: Regression Binary Variables

Correlation, Formula 0 2 4 6 8 10 12 14 16 18 0 1 2 3 4 5 6 7 8 9 Positive realtionship 0 2 4 6 8 10 12 14 16 18 0 1 2 3 4 5 6 7 8 9 Negative Relationship

r

=

n

(

å

xy

)

-

( )

å

x

( )

å

y

n

å

x

2

-

( )

å

_x

2

é

ëê

ù

ûú

n

y

2

( )

å

y

2

å

é

ëê

ù

ûú

(27)

The correlation test for this study was executed in Microsoft Excel, only one variable showed a meaningful correlation, the other variables were re-expressed in an attempt to straighten the data, by taking the log of x and y but also by taking the square root of x. However, the correlation did not improve, an example is presented in Figure 12.

To perform a

regression that gives any useful information the data needs to follow the straight enough condition, if the data is behaving like in Figure 12, the data isn’t straight enough and the regression analysis won’t say anything about the relationship of the variables (De Veaux et al. 2014:250).

The r-squared value is the correlation times itself and is therefore a number between 0-1 and gives the percentage of how well the regression line fits the data. If the data is scattered close to the regression line the coefficient will be close to 1 and if the data is scattered far from the line it will get closer to 0. The aim for using regression analysis is to investigate how much of the change in the y-variable is explained by the x-variable. Although the percentage of the change in the dependent variable that is explained by the independent variable is high it still can’t conclude causality, other factors outside the regression model might affect the dependent variable. The probability that the regression reflects the entire population is given by the p-value. For this study the 5% level is used, meaning that if the p-value is lower than 0,05 the results are 95% statistically certain they reflect the population (De Veaux et al. 2014:213, 534, 709, 713).

Outliers in the data are observations that are considerably different from the rest of the data; their values are substantially larger or smaller than the other data. The outliers need to be considered carefully when performing any statistical tests since their extreme values can have a powerful effect on the outcome of the analysis. Removing the outliers altogether won’t necessarily give a more accurate view of the issue; therefore it is important to perform statistical tests with and without outliers to fully understand their impact on the outcome (De Veaux et al. 2014:250). Therefore the correlation and regression analyses were performed on data with and without the outliers. y = 0,2223x + 0,1948 R² = 0,1399 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5 0 1 2 3 4 5 6 7 Number of Pictures

(28)

Four regression models were constructed for this study; each model was calculated with and without outliers. The outliers were identified through calculation the

Interquartile range and constructing boxplots for each variable. Projects that received over 500% of their funding goal were outliers, removing these resulted in different effects on correlation and r-square value for the models.

Model A consists of only one variable, “comments”, since it was the only continuous variable that had an adequate correlation to the dependent variable. Model B Consists of only two binary variables the most common variables from the Visual category. Four variables make up Model C, the most common variables from both Visual and Expertise. Model D consists of the most common binary variables from Expertise and Visual and “comments”.

The models were calculated in Microsoft Excel and follow the regression formula. Where B0 is the intercept where the regression line cuts in the y-axis, B1 is the slope of the line, E stands for the error term and y is the expected value given the regression equation. The regression equations used for the different models is presented in Table 6.

Criticism

Firstly the definition and measurement of quality and the “quirky element” demands attention in this section. There is a certain level of subjectivity regarding the

definition of these variables, which affects both the study’s construct validity and reliability (Bryman and Bell 2005:93-96). These variables are still interesting since there are clear differences in quality of the media published on the campaigns. However the definition of quality is, because of the nature of the term, subjective. Transparency in the analysing process is adopted to mitigate these inconsistencies, and also attempts to clarify and visualise the definitions of the variables. As for the “quirky element” which is defined by humour is suffering a great risk of researcher

Table 5, Regression Models

(29)

Moreover the validity of the research suffers from low generalizability (Bryman and Bell 2005:100-101), although its quantitative approach the sample of 150 campaigns is not large enough to accurately generalise the result for the entire population. In regards of the construct validity, as mentioned above, is affected by subjectivity in the selection and definition of some of the variables. However the study offers altogether 19 independent variables that aim to thoroughly measure the issue and the majority of these are not suffering from a subjective biases.

Additionally the quantitative characteristics of this study result in it not describing the phenomenon and its causes sufficiently. To gather an in-depth understanding of this problem, qualitative studies such as interviews with creators and funders could be appropriate.

The sample was not picked randomly, not all the campaigns in the studied population had the same chance of being selected. To achieve a random sample all reward-based crowdfunding platforms should have been part of the sampling process, however this would make it more difficult to compare the projects since the platforms are designed differently.

The reliability of the study, also affected by the subjectivity in the definitions of quality, most likely suffers more than the validity since the reliability concerns the researchers impact on the study (Bryman and Bell 2005:94). However the research questions demands a definition of these subjective elements, therefore transparency is crucial to understand the results properly.

The transparency in methods also eases the possibility of replicating this study by another researcher.

There is also the issue whether the measurements developed to study this problem are accurate or not. It is possible that other variables would offer a greater insight in these issues, however the development of the variables are based on the outlook set by earlier research as well as a thorough review of the platform.

Source Criticism

The majority of the sources used in this study consist of research papers, scientific articles and textbooks that are recognised and established. The research articles and other secondary sources were created in another purpose than this study this has been considered throughout the study. There are also articles from web-based magazines like Forbes’ online magazine, and other online sources. These online sources are due to crowdfunding’s nature reasonable to use online since it’s an online phenomenon and a lot of information is impossible to find elsewhere.

(30)

Results

The results of the study are presented by first summarising general tendencies of the data. Then the binary and continuous variables are then presented separately and are followed by an analysis of the most significant variables combined together.

Overview

The successful campaigns generally utilise all variables to a larger extent than the campaigns that failed. The sample consists of campaigns from all continents (except Antarctica and the Arctic) of the world but the majority of the campaigns origin from the United States followed by campaigns from Europe. The successful campaigns have more quality in their videos. The failed campaigns don’t outperform the successful campaigns for any of the variables. The most common variables are the same for the successful as the failed campaigns however they are ranked differently, where the visual variables are more common for the successful and the expertise variables are more common for the failed campaigns. Moreover one variable that wasn’t studied specifically was in what manner the expertise variables were

presented, but it is noteworthy that the timeframe and budget often were presented by a graph, timeline or picture rather than by text. The “quirky element” variable was often utilised in the video, for instance by having bloopers at the end of it.

Binary Variables

The data shows that campaigns that don’t communicate their business plan are not successful in reaching their goal. Only one project in the sample was successful in receiving funding without communicating any of the business plan-variables. The business plan variables showed different frequency in the investigated campaigns, both for the successful and failed projects. However the successful campaigns have overall scored higher for all variables than the failed campaigns. The least common expertise-variable was “budget”, which was only mentioned in 28% of the successful campaigns and 20% of the failed campaigns, and 24% for the total sample. The most common of the binary variables for the successful campaigns was “video”, followed by “picture quality” and third most common was “risks”. For

Origin of Campaigns Asia Africa Austrlia Europe North America South America

(31)

the failed campaigns “risks” was the most common variable and “video” came second, followed by “picture quality”.

As for the variables that showed the greatest difference between successful and failed projects are presented in Table 8. “Video quality”, “picture of creator” and creator experience” are both amongst the most common variables and the greatest difference. Although they are most common for both successful and failed campaigns they are also showing big difference between successful and failed projects.

The most common variables that concern the business plan were “risks”, “creator experience” and “timeframe”, as mentioned the budget isn’t commonly

communicated and the strategy for mitigating the risks is mentioned in 33% of the successful respectively 32% of the failed campaigns. It is common to mention what risks a project could entail but less common to offer a strategy that will mitigate these risks. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Yes/Quality/Mentioned % Successful 1 Failed 1

Figure 14, All Variables

(32)

The variables that aim to measure the creator’s likeability and honesty, collectively referred to, as the “trustworthiness variables” are generally less common than the other variable-groups. The most common element of these was the “picture of creator”-variable; followed by “mention another actor”, 53% of the successful campaigns have mentioned an external actor. The failed campaigns however have more commonly used storytelling although still in a smaller extent than the successful campaigns. Generally in the campaigns the product or service were described and the creators background were left out. Another uncommon element was the combination of all the Trustworthiness and Visual categories, which was only found in three campaigns that all reached their funding goal.

80% 33% 64% 28% 75% 76% 32% 43% 20% 51% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Risks Strategy Timeframe Budget Creator exp.

Expertise/Business Plan Variables (Yes/Mentioned)

Successful % Failed %

Figure 15, Expertise Category

60% 31% 53% 25% 29% 25% 17% 5% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Pic of Creator Storytelling Mention an. Actor Quirky Element Trustworthiness Variables (Yes/Mentioned) Successful % Failed %

(33)

The Visual variables score the highest across successful and failed campaigns, with the “Video Quality” for failed campaigns being almost half of the successful. Of the successful campaigns 93% had a video, 79% of those videos qualified as quality videos and 85% of the campaigns had quality pictures. For the failed campaigns 71% had a video, 40% of these were of quality and 55% of the campaigns had quality pictures.

The variables that showed great differences between successful and failed projects were tested through chi-squared test to ensure statistical significance for the difference.

The H0 assumption is that the variables are independent and do not show a relationship between successful and failed campaigns for the different variables tested, meaning that the two variables do not have a connection. The H1 says the opposite that there is a difference between the successful and failed campaigns that the variables are dependent and have a connection. The results are shown in the table below, where the H0 assumption is rejected for variables “mention another actor”, “video quality” and “Kickstarter loves”. This means that statistically there is

difference between success and failure for the variables “picture of creator”, “creator experience” and “quirky element”. However these tests says nothing of how these variables affect the dependent variables success or failure, only that there is a difference between the two.

93% 79% 85% 71% 40% 55% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Video Video Quality Picture Quality

Visual Variables (Yes/Quality)

Successful % Failed %

Figure 17, Visual Category

Dazzled to invest - Are the funding masses falling for effects?

Av: Jessica Österberg